PEI (parental education involvement) encompasses various behaviors that parents adopt to enhance students’ educational achievements based on diverse educational concepts and academic achievement expectations (Seginer, 2006). The positive impact of PEI on middle-school students’ learning engagement has been consistently emphasized in research (Chen et al., 2024). Active PEI correlates with enhanced academic outcomes, increased attendance rates, and overall student success. When parents engage in their children’s schooling, students demonstrate higher academic achievement, school engagement, and motivation (Barger et al., 2019). Studies among 9th and 10th graders in Jordan showed that PEI positively influenced students’ emotional engagement in school (Sapungan and Sapungan, 2014). This implies that students with more involved parents are more likely to find school enjoyable, have high self-esteem, and perceive school as a satisfying experience. Thus, parental support and guidance play a crucial role in positively affecting students’ learning.

Few studies have directly explored the relationship between GEI (grandparental education involvement) and learning engagement. However, current research suggests that grandparental involvement has a positive impact on children’s social skills and reduces the occurrence of internalizing and externalizing problem behaviors, which may subsequently influence adolescents’ learning engagement (Luo et al., 2020). Yorgason et al. (2011) investigated the association between grandparental involvement and positive outcomes for grandchildren. They found that, when grandparents are involved in their grandchildren’s daily lives, the grandchildren exhibit greater sociability and increased learning engagement. Therefore, we propose the following hypothesis:

PGI encompasses cognitive and behavioral tendencies focused on active and intentional personal growth. This concept has been delineated into four relatively independent dimensions: readiness for change, planfulness, using resources, and intentional behaviors (Dengyong et al., 2014). According to the theory of self-determination, individuals generally have three basic needs: autonomy, ability, and relationships (Deci and Ryan, 2000). PGI is a manifestation of the need for autonomy as a vital intrinsic motivation that can significantly influence academic performance and emotional adaptation (Grolnick and Slowiaczek, 1994). Researchers have suggested that positive academic performance is challenging unless students autonomously engage in learning (Schnitzler et al., 2021). When students’ autonomous needs are met, they typically experience a sense of choice and self-recognition of their behavior (Deci and Ryan, 2008; Gagné and Deci, 2005). Consequently, they become more willing to participate and invest in learning activities. Studies have indicated that fulfilling students’ autonomy needs enhances their internal motivation, leading to more positive learning engagement behaviors (Reeve, 2012). Several studies have explored the relationship between PEI and students’ PGI, particularly among Chinese primary school students (Zhao et al., 2021). The findings reveal that the manner and extent of PEI positively impact students’ autonomy. Parent’s support and encouragement are crucial in cultivating students’ PGI, which closely correlates with their learning engagement (Li et al., 2023). In summary, PEI and GEI are advantageous for students’ PGI, and PGI is linked to improved learning engagement. This suggests that PGI may mediate the relationship between PEI/GEI and learning engagement. Thus, we propose the following hypothesis:

Apart from PGI, an individual’s self-education expectation also modulates one’s implementation of learning activities. Students’ educational expectations are closely related to their academic performance (Pinquart and Ebeling, 2020). Students with higher self-education expectations work harder to learn to meet their expectations and achieve relatively better academic performance. One study proved that middle-school students’ self-education expectations has a positive impact on their academic development and can promote their degree of investment (Li et al., 2020; Shi et al., 2023). According to Bandura’s triadic interactive determinism, behavior, cognition, and the environment are interconnected and interact during the social learning process (Chai and Ye, 2024). Furthermore, the external environment influences individuals through inter-subjective factors. In this study, PEI and GEI were considered external environmental factors, while self-education expectation was considered an internal factor affecting individuals. Based on Bandura’s ternary interaction theory, junior-high-school students’ learning engagement is influenced not only by PEI, GEI, and self-education expectation, but also by their combined effect (Schmid and Garrels, 2021). Therefore, self-education expectation may act as a moderator in the pathway between PEI/GEI and learning engagement. Thus, we propose two hypotheses: self-education expectation significantly moderates the relationship between PEI and learning engagement (H3a), and self-education expectation significantly moderates the relationship between GEI and learning engagement (H3b).

The present study examined the mediating and moderating factors in the association between PEI/GEI and learning engagement among adolescents in the Chinese context. It also explored the underlying mechanisms of this relationship by considering cognitive trait factors, including the mediating role of PGI and the moderating role of self-education expectation. By constructing two moderated mediation models (see Figure 1), this study contributes to the understanding of the mechanisms and differences underlying the influence of PEI and GEI on adolescents’ learning engagement. Additionally, it provides empirical support to guide adolescents’ learning engagement in a scientifically practical manner.

This study was approved by the Ethics Committee for Scientific Research at the authors’ institution. Data were collected from 850 junior-high-school students using a paper-based questionnaire. After excluding invalid responses, 822 responses were retained, resulting in a response rate of 96.7%. Among these valid responses, 380 were from male students and 442 were from female students, with ages ranging from 12 to 15 years (mean ± standard deviation: 12.16 ± 1.20). The participants were distributed across different grades, with 522 (63.5%) in Grade 1 of junior high school and 300 (36.4%) in Grade 2. Additionally, we categorized students based on their main caregivers, with 636 (77.4%) raised by parents and 186 (22.6%) raised by grandparents.

First, we conducted descriptive statistics and correlation analyses using IBM SPSS 26.0. Subsequently, we tested the mediation and moderated mediation models using the IBM SPSS macro PROCESS (Hayes, 2013), which has been widely used to test complex models, including mediated moderation and moderated mediation models. Models 4 and 5 in the PROCESS macro program¹ were used to analyze the mediating role of PGI and the moderating role of self-education expectation between PEI/GEI and learning engagement. Additionally, previous research has revealed that learning engagement differs based on grade and previous achievement (Piñeiro et al., 2019; Zhou, 2021). Therefore, grade and previous achievement were included as control variables in this study.

The results in Tables 1, 2 illustrate the means (M), standard deviations (SD), and correlation coefficients of all the variables. Not all forms of parental involvement are positively related to academic achievement (Boonk et al., 2018). It is necessary to explore the different dimensions of parental/grandparental educational involvement. The four dimensions of PEI (r = 0.113–0.731, p < 0.01) and the four dimensions of GEI (r = 0.146–0.629, p < 0.05) demonstrated significant positive correlations with learning engagement, PGI, and self-education expectation.

Separate multiple regression analyses were conducted using the four dimensions of PEI and GEI as independent variables. As shown in Tables 3, 4, we found that, after controlling for grade and prior academic achievement, the PEI results indicated that emotional leisure (β = 0.333, p < 0.001) had the most decisive influence, whereas the GEI results showed that academic support (β = 0.277, p < 0.01) was most influential.

We employed Model 4 of PROCESS to examine the mediating effect of PGI between PEI/GEI and learning engagement. Percentile bootstrapping and bias-corrected percentile bootstrapping with 5,000 resamples were employed to construct 95% confidence intervals for the indirect effects. The four dimensions of parental/grandparental educational involvement represent different aspects of educational involvement, and they showed different correlations with learning engagement. Therefore, we will explore the relationship between educational involvement and learning engagement by dimension. After controlling for grade and previous achievement, four dimensions of parental education involvement significantly and positively predicted learning engagement in the absence of the mediator. Among them, the prediction power of emotional leisure is the strongest. Therefore, we mainly present the results of emotional leisure, as shown in Table 5.

As observed in Tables 5, EL was positively correlated with PGI (β = 0.422, p < 0.001). When EL was controlled for, PGI was positively correlated with learning engagement (β = 0.421, p < 0.001). Moreover, when PGI was included, the association between EL and learning engagement remained significant (β = 0.213, p < 0.001). Finally, a bias-corrected percentile bootstrap method was used to test the mediation model. We generated 5,000 bootstrapping samples from the original data (n = 636 from parent data) by random sampling. The results (Table 6) indicate that the indirect effect of PGI was 0.269, and its 95% CI was [0.192, 0.350] in the parent data. This mediating effect accounted for 45.516% of the total effect of the association between ER and learning engagement. In other words, PGI partially mediated the association between EL and learning engagement.

Four dimensions of grandparental education involvement significantly and positively predicted learning engagement in the absence of the mediator, thus supporting H1. Among them, the prediction power of academic support (AS) is the strongest. Therefore, we mainly present the results of academic support, as shown in Table 7.

As observed in Table 7, AS was positively correlated with PGI (β = 0.434, p < 0.001). When AS was controlled for, PGI was positively correlated with learning engagement (β = 0.466, p < 0.001). Moreover, when PGI was included, the association between AS and learning engagement remained significant (β = 0.222, p < 0.01). Finally, a bias-corrected percentile bootstrap method was used to test the mediation model. We generated 5,000 bootstrapping samples from the original data (n = 186 from grandparent data) by random sampling. The results (Table 8) indicate that the indirect effect of PGI was 0.263, and its 95% CI was [0.164, 0.364] in the parent data. This mediating effect accounted for 47.645% of the total effect of the association between AS and learning engagement. In other words, PGI partially mediated the association between AS and learning engagement. Thus, H2 was supported.

We used Model 5 of PROCESS to determine whether the direct effect was moderated by self-education expectation. We introduced an interaction effect between self-education expectation and education involvement (emotional leisure, teaching rules, academic support, life care) to predict learning engagement. The unstandardized model estimates for H3a and H3b are presented in Tables 9, 10, respectively. Due to space limitations, only the data results related to emotional leisure have been presented.

For the situation where parents directly raise their children, the interaction term (emotional leisure/teaching rules/academic support × self-education expectation) significantly predicted learning engagement, indicating that self-education expectation moderated the direct effect of emotional leisure, teaching rules, academic support on learning engagement, thereby supporting H3a partially. The positive direct effect of emotional leisure, academic support on learning engagement was significant for individuals with low and high self-education expectation. However, the same positive direct effect of teaching rules was significant only for individuals with low self-education expectation. To better understand the moderating effect of self-education expectation, plots of the relationship between emotional leisure, academic support, teaching rules and learning engagement, at two levels of self-education expectation (1 SD below the mean and 1 SD above the mean) are depicted in Figures 2–4, respectively.

For the situation where grandparents providing caregiving for grandchildren, only the interaction term emotional leisure × self-education expectation significantly predicted learning engagement, indicating that self-education expectation moderated the direct effect of emotional leisure on learning engagement, thereby supporting H3a partially. Meanwhile, the positive direct effect of emotional leisure was significant only for individuals with low self-education expectation. To better understand the moderating effect of self-education expectation, plot of the relationship between emotional leisure and learning engagement, at two levels of self-education expectation (1 SD below the mean and 1 SD above the mean) is depicted in Figure 5.

As observed in Table 10 and Figure 5, for individuals with low self-education expectation, emotional leisure was positively associated with learning engagement (conditional effect = 0.398, SE = 0.127, 95% CI = [0.220, 0.796]). For individuals with high self-education expectation, the direct effect between emotional leisure and learning engagement was not significant (95% CI = [−0.349, 0.098]).

We observed that both PEI and GEI positively predicted adolescents’ learning engagement, which aligns with existing theoretical perspectives and empirical evidence. Positive parenting styles can directly influence adolescents’ psychological development (Bi et al., 2024). A positive association between PEI/GEI and heightened learning engagement levels ultimately predicts elevated academic achievement (Day and Dotterer, 2018; Hoover-Dempsey and Sandler, 1995; Skinner et al., 2009). This implies that individuals who experience greater PEI/GEI are more inclined to engage in learning activities. Drawing on social capital theory, family, as a significant social network, expands students’ social capital, providing more learning opportunities and resources. Meanwhile, four dimensions of parental education involvement significantly and positively predicted learning engagement in the absence of the mediator. In parents’ data, the prediction power of emotional leisure is the strongest. Compared with grandparents, parents are stricter with their children. Therefore, children need more leisure activities when they are with their parents. Positive emotional experiences from nature are of significance (Wang et al., 2025). Leisure activities positively predicted positive emotion (Zhang and Zheng, 2017) which helpful for learning engagement. In grandparents’ data, the prediction power of academic support is the strongest. Grandparents can, to a certain extent, make up for the lack of educational support caused by the absence of parents by means such as supervising homework and helping solve learning difficulties, thus enabling children to devote themselves to their studies.

Our study not only revealed that PEI, GEI, and PGI positively predict learning engagement, but also unveiled the mediating role of PGI in the relationships between PEI/GEI and learning engagement. This finding aligns with previous studies supporting the idea that students who experience PEI/GEI are more likely to value their learning, thus promoting autonomous motivation (Schmid and Garrels, 2021). According to Deci and Ryan’s self-determination theory (Vansteenkiste et al., 2020), individuals have three basic psychological needs: autonomy, relatedness, and competence. PEI and GEI play a crucial role in shaping children’s future development, fostering autonomy, and motivating active participation in self-regulated learning (Ahmed et al., 2024). When adolescents’ basic needs are satisfied, they are more motivated to engage in their growth process and have higher expectations for their future academic life (Leversen et al., 2012). Parents and grandparents who are positively involved in offering academic assistance or life care satisfy students’ psychological needs and foster their autonomous motivation for learning. As a vital component of an individual’s inner psychological resources, PGI plays a mediating role in the connections between PEI/GEI and adolescents’ learning engagement. A supportive family atmosphere fosters stronger autonomous motivation, and adolescents will have higher expectations for learning when they are effectively assisted in developing positive character traits such as PGI (Benneker et al., 2023). Furthermore, adolescents who grow up in such environments often have more positive experiences, experience better psychological wellbeing, and exhibit greater life satisfaction (Wang et al., 2018). Consequently, they are more likely to enjoy their academic lives, which in turn influences their learning engagement.

Our study revealed that adolescents’ self-education expectation moderated the relationships between PEI/GEI and learning engagement. In the data of the parents, the interaction term (emotional leisure/teaching rules/academic support × self-education expectation) significantly predicted learning engagement, indicating that self-education expectation moderated the direct effect of emotional leisure, teaching rules, academic support on learning engagement. In the data of grandparents, only the interaction term emotional leisure × self-education expectation significantly predicted learning engagement, indicating that self-education expectation moderated the direct effect of emotional leisure on learning engagement.

Specifically, these relationships were stronger among adolescents with lower self-education expectations than among those with higher expectations. According to existing research, adolescents with higher academic achievement expectations are more likely to prioritize personal effort and autonomous motivation (Shengyao et al., 2024). Conversely, when adolescents lack aspirations for improved academic achievement, PEI/GEI can instill a sense of responsibility and motivation, driven by the desire not to disappoint their parents/grandparents, thereby empowering them to tackle challenging learning tasks more effectively. In this context, PEI/GEI serves as a source of stability and reassurance, assisting adolescents in navigating diverse obstacles. Such experiences can enhance adolescents’ awareness of the support they receive from others, ultimately promoting their learning engagement (Luo et al., 2020).

However, the moderating mechanisms of self-education expectation may differ depending on adolescents’ perceived PEI/GEI. In the data of the parents, for individuals with low self-education expectation and high self-education expectation, the positive direct effects of both emotional leisure and academic support on learning engagement were significant. However, the positive direct effect of teaching rules on learning engagement was not significant for individuals with high self-education expectation. This means that, for individuals with high self-education expectations, teaching rules did not affect learning engagement. In the data of the grandparents, self-education expectation only moderated the direct effect of emotional leisure on learning engagement. For individuals with low self-education expectation, emotional leisure was positively associated with learning engagement. For individuals with high self-education expectation, the direct effect between emotional leisure and learning engagement was not significant. In China, the ways and values of parents and grandparents differ significantly (Lu et al., 2022). Parents were more influenced by modernization and urbanization, with a greater emphasis on academic performance, personal achievement, and social status. Thus, they may be more inclined to use modern educational resources and methods, such as cram schools and online education, in the hope that their children will stand out from the fierce competition (Chen et al., 2021). According to existing research, children’s learning engagement is positively related to both paternal and maternal involvement (Li et al., 2023). However, grandparents tend to pay more attention to the inheritance of traditional values (Zhang and Wu, 2021) or children’s health outcomes (Pulgaron et al., 2016). Thus, grandparents’ influence might differ owing to generational gaps and differing perspectives on education (Grolnick and Pomerantz, 2022). While grandparents may still play a role, their influence may be indirect or less pronounced than that of parents (Dunifon and Bajracharya, 2012).

This study has important theoretical and practical implications. First, it aligns with the theory of positive adolescent development, which emphasizes how adolescents’ external environments interact with their inner strengths to shape their learning engagement. This offers effective strategies for fostering adolescents’ learning behaviors. Second, there is an urgent need to prioritize the promotion of autonomous motivation and initiative among adolescents. This can be achieved by clarifying their learning goals and fostering personal growth dynamics. These qualities play a pivotal role in adolescents’ learning behaviors and growth processes, equipping them with resilience to overcome challenges and fulfill their potential.

Regarding practical implications, this study considers the discrepancy mechanisms in which PEI/GEI affects students’ learning engagement. The situation where parents raise their children themselves, along with taking them to educational places like science museums, engaging in leisure activities such as traveling or exercising, can help enhance children’s learning engagement. When parents work in other places and the children are raised by their grandparents, the grandparents providing necessary academic support for the children can help increase their learning engagement. However, for the elderly with lower educational levels, this is very difficult to achieve. Meanwhile, the results of the moderation effect revealed that parents play a more important role in contributing to children’s learning.

Despite the valuable insights provided by this study, it has some limitations. First, the study’s scope was confined to one secondary school in China, potentially introducing regional bias and limiting the generalizability of the results. To enhance the reliability and applicability of our findings, future research should expand the sample size and consider diverse cultural contexts. Second, longitudinal studies and follow-up surveys could offer a more dynamic and holistic comprehension of PEI and adolescents’ learning engagement; thus, longitudinal studies should be conducted in the future.